AGS-Imager; 2-axis Scan Mirror, Deployable Calibrators

This conceptual layout was for an early version of the design that was completed to see if the windows and filters could be installed and configured to correct of astigmatism caused by having "planar" elements located in converging light. The design concept was successful in eliminating astigmatism. In the 3D version actually implemented, the VIS/SW arm was rotated 90-degrees by rotating the main Beam Splitter 90-degrees. Also, a fold mirror was located between the Pupil Stop and the VIS/SW Entrance Window. The Fold Mirror allowed the overall instrument to fit within the space available. The use of the Fold Mirror reduces the system weight by allowing a smaller instrument optical bench and enclosure.

[A] 2D conceptual layout of ZEMAX Y-fan raytrace of 1-degree diameter field-of-view.
[B] Close-up LWIR/MWIR Vacuum Dewar and SWL/VIS Vacuum Enclosure.

Optical ray traces of all four bands were completed with an edge ring of rays spaced every 22.5 degrees around the Telescope Entrance Pupil. This set of rays defined the full diameter of the light beams and clearly located the field and pupil stops. By convention, the coordinate system was rotated so that the main beam from the telescope was horizontal and the scan mirror directed the field-of-view vertically upward from the earth viewing face. for this conceptual configuration, the polar face (never illuminated by sunlight) is the face to the left behind the Primary and Tertiary Mirrors. This is approximately a 180-degree rotation from the conceptual optical layout shown above. The 3-D layout was completed in AutoCAD, Release 14. The optical rays were enclosed in transparent solids so they could be seen in photo-realistic renditions of the structure (the renditions cause individual lines of zero width to disappear). A "zero" coefficient of thermal expansion, graphite fiber structure was put in place to hold all the system components in position.

[D & E] Orthographic Side View of the AGS Imager - Photo-Realistic and Hidden-Line renditions showing the overall light path from the earth to the detector enclosures. Many instrument parts are turned off or made transparent so that the full conceptual 3-D layout can be visualized.
== Starting from the top, a deployed sun shield is shown. The light path from the earth (cyan) enters from the top and strikes the two-axis scan mirror. A 2-axis scan mirror was used to make the scanning system more repeatable and higher precision. The north-south scanner axis is perpendicular to the side view. It increments in 0.4-degree steps so that the east west scans are separated by approximately 0.8-degree intervals on the earth. The east-west scan axis is parallel the the scan mirror face (lines locating these axes can be seen in the appropriate hidden-line renditions). The Scan Mirror shown is identical in outline to the GOES Imager Scan Mirror for GOES 8-10.
== Light reflects to the left from the Scan Mirror and enters a well baffled telescope. The Primary Mirror is mounted to a vertical structure that extends straight down and also provides the mounting structure for the Tertiary Mirror located directly below. The Scan Mirror and Primary Mirror are located above and attached directly to a rectangular box-like set of tubes which form the main instrument optical bench.
== Light is reflected from the Primary Mirror and proceeds toward the Secondary Mirror, to the right and downward. Only part of the Secondary Mirror can be seen below the upper tubes forming the optical bench box. The red mount for the base of the Secondary is easily seen in the larger image.
== A Field Stop is located just to the left of the Secondary Mirror. Baffles completely separate the Telescope from the Optical Bench, with the only entrance permitted through the Field Stop. The detector electronics box (blue) is located on the sidewall behind the Field Stop and Detector Enclosures. Baffles around the Field Stop volume can be seen in the hidden-line and isometric views.
== The Tertiary Mirror reimages the image formed at the field stop onto the detectors within vacuum enclosures. The band separation is achieved using dichroic beamsplitters which reflect the shortest wavelengths in all cases for this design. The vacuum enclosures are used to permit operation in air on the ground. The VIS/SW focal plane is passively cooled to optical bench temperature to minimize electronic noise. A thermo-electric cooler might be required for these two detectors. The MW/SW enclosure is a cryogenic dewar used to cool the detectors and local environment up to and including the Pupil Stop so that the background radiation does not swamp the signal from Earth.

[F & G] Orthographic End View of the AGS Imager from Behind the Scan Mirror - Photo-Realistic and Hidden-Line renditions. Lines corresponding with the Scan Mirror axes can be seen in the hidden-line image. The VIS/SW detector enclosure is the (yellow) cylinder to the left on the bottom of the optical bench box. The MW/LW detector enclosure (red) is located on the Instrument central plane. A vertical (blue) box located to the right of the optical bench contains the detector electronics. The exact size of the box is not critical to this design because it could be twice as thick and extend nearly the full length of the optical bench without interfering with the light paths. It could also be relocated underneath the optical bench.

[H & I] Orthographic Bottom View of the AGS Imager - Photo-Realistic and Hidden-Line renditions. L

These views show the AGS Imager from the bottom. In the Telescope and Optical Bench base baffles and bench faces have been turned off so the optical path can be seen. The back and base of the Scan Mirror Assembly can be seen in [H], but were not shown in the hidden-line version [I]. The location of the Field Stop is marked by a (grayish) vertical bar behind the center of the MW/LW (red) detector electronics in [H] which is a tilted baffle plane. It is nearly impossible to distinguish the Tertiary Mirror from the Primary Mirror except by following the edges of the light beams carefully. The Tertiary Mirror is about half the diameter of the Primary.
== The "U" shape of the Sun Shield is apparent in this view. The side that corresponds with the side used for radiative cooling is "omitted" because it is never needed. As well a not being a scattering source, the "omitted" side does not heat up and re-radiate into the Imager entrance aperture. In order to keep sunlight from illuminating the radiative cooler face, the instrument is rotated 180-degrees around the earth-looking axis at the solar equinoxes.

[J] Close-up of the Detector Enclosures from the Bottom - Photo-Realistic rendition.

The common light paths (cyan) that can be seen in this bottom view are the outermost edges of the main path (nearly cylindrical) into the Telescope and the reflected (converging from the left) path from the Primary Mirror toward the Secondary Mirror. The central parts of these paths are blocked by the dark base baffle which separates the Telescope from the Optical Bench Box. To the left are seen the diverging and converging common paths (cyan) from the Field Stop to the Tertiary and from the Tertiary to the Main Beam Splitter. The Main Beam Splitter housing (gray with green mount) holds the beam splitter that transmits the MW/LW light and reflects the VIS/SW light 90-degrees (in this view). The long snout on the MW/LW cryogenic Detector Enclosure (red) contains a window in front of the Pupil Stop, allowing the Pupil Stop to be cooled to the detector temperature. The location of the Pupil Stop in the VIS/SW path is where the path shows a slight change in slope, about half way between the Main Beam Splitter and the Fold Mirror. That Pupil Stop does not need to be cooled, but is used for stray light control only. There is a window at the entrance to that detector enclosure. The Detector Electronics Box (blue) is just below MW/LW Detector Enclosure.

[K] Hidden-line Isometric View - Because a 2-Axis Scan Mirror is chosen for pointing stability and control, it is necessary to have deployable calibrators. Because the interface volume is not well defined (we don't know where deployables can extend outside the volume limits during operation) this configuration has selected a linear actuated calibrator mounted on the top of the Imager enclosure. There is plenty of room for this device. It is assumed that the Thermal Calibrator will deploy approximately once every thirty minutes. The Perforated Plate Calibrator will be used for a few minutes, a few times around midnight, at the spring and fall equinoxes. In this design the same actuator deploys both calibrators. In these views [K, L, M, N] the Perforated Plate Calibrator is in the "calibrate" position and the Thermal Calibrator is stowed. Rotary actuators would be mechanically more preferable, but the interface is not well enough defined to use them for this conceptual design. The Sun Shield is also required to be deployable, but it is a one-time operation at the beginning of the mission. Again, because we don't have enough definition of the interface, the concept is to have the Sun Shield stowed over the entrance port during launch. It is a articulated deployable. It is not clear if the Sun Shield will be permitted in the position shown because there may be interference with other instruments or Spacecraft Star Trackers. The Sun Shield improves performance around midnight by decreasing the heat load on the Scan Mirror and into the Entrance Port, it is not required to make operations possible.

[M] Isometric Rendition of the Full Imager. - Photo-Realistic Rendition - See discussion under [K]

[N] Closer View of the Isometric Rendition. - Photo-Realistic Rendition - See discussion under [K]

[O] Side View of the Detector Enclosure/Secondary Mirror/Field Stop - The thin baffles seen edge-on are difficult to see in this view.

[P] Isometric View of the Detector Enclosure/Secondary Mirror/Field Stop - The baffles are easier to see in this view. There is a baffle box completely surrounding the Detector Enclosures, except where the light path enters. This box volume will be cooled to about 200K by coupling to the radiative cooler. The MW/LW (red) Detector Enclosure will be further cooled using a mechanical cooler.

[Q] Isometric View of the Detector Enclosure/Secondary Mirror/Field Stop - This hidden-line rendition shows the baffle edges more distinctly.

[R] For the strawman design illustrated above, the concept is for Silicon Carbide Mirrors to be mounted to a Graphite Epoxy Optical Bench. The structure is rather simple, being suggested as rectangular structures with stiffening face sheets. Support tubes have been added at load points, such as the Detector Assemblies. This simple structure is adequate for estimating volumes, weights, and thermal deformations (STOCS Analysis). Tailor-made, ultra-lightweight structures are possible when the design is optimized in detail. An example of an ultra-lightweight structure is shown in Figure [R]. The sketch is based on a SiC structure as well as SiC mirrors which have been fabricated and tested for an IR system. The details and performance of the structure are reported in SPIE Volume 2543 (1995) Silicon Carbide Materials for Optics and Precision Structures. Other, integrated structures are also shown in that document. These structures are, at first glance, somewhat similar to the AGS Imager so similar structures are likely to work for the Imager. There are enough differences (scan mirror, stray light baffling, solar intrusion, heavier detector assemblies, etc.) that the simple triangular structure may not work. It looks like a triangular cross section lower assembly could be used. The upper surface that supports the telescope and scan mirror would be the base of the triangle, with the apex forming an axis where the detector assemblies could be mounted. This might reduce the structure weight by 20 to 25 percent. The resulting structure would look like the Hubble Space Telescope instrument graphite epoxy structures.

The GOES-R specification requires 80% of the encircled energy to be within a 1-km circle for the VIS bands and within a 4-km circle for all the other bands. The GOES "super"-R specifications are to reduce those diameters by a factor of 2, i.e 1/2-km for the VIS and 2-km for the other bands. The present design performs better than the GOES-R specification for all bands and is better than the GOES "super"-R specifications, except for the LW band which is beyond the diffraction capability for a 30-cm diameter telescope. It would require increasing the diameter of the telescope to 40+ centimeters to achieve the performance in the LW band. The corresponding overall size and weight penalty will probably not be accepted and the GOES "super"-R LW performance specification reduced. A similar reduction in performance specification for this same long wavelength range has been implemented in the GOES IJK/LM specification. The GOES IJK/LM specification is based on Modulation Transfer Function (MTF) values and not encircled energy. For comparison, the AGS Imager performance is presented in MTF form.

The Beam Splitters are wedged to eliminate astigmatism introduced by using them in converging light. Performance is optimized at 0.55, 2.0, 4.0, and 10.0 microns wavelength. The Telescope optical prescription is optimized so the geometric spots sizes are optimum in the VIS channel. The VIS channel has 0.012-mm pixels. System performance is limited by detector size in VIS band, and dominated by diffraction in all other bands. Performance has been analyzed using ZEMAX. The enclosure windows and Focal Plane Assembly filters have been included in the analysis.

Encircled Energy Curves - The encircled energy curves are marked to show the 80% performance requirement and where it is met for a perfect diffraction limited telescope. The minimum and maximum radii for the 12 field positions are shown. It looks possible to locate the detectors to avoid the worst case condition (even though that worst case condition generally still meets specification). The encircled energy specification is the equivalent of specifying the MTF performance at the Nyquist limit, without giving any other characteristics of the shape. In this sense, the specification is somewhat more stringent than the MTF specifications transferred from GOES-R.
... The VIS band is dominated by geometric performance and is most sensitive to slight changes in performance specification. The other bands are all dominated by diffraction and the 80% specification value corresponds to the dimension where the Airy diameter is at or past the edge of the detector pixels. A small change in the performance specification causes large changes in the encircled radii for the SWIR and MWIR bands. If the encircled energy specification were 70%, the performance of the SWIR and MWIR bands could be considered essentially diffraction limited. For the LWIR band, the radius that just fits within the detector pixel contains only about 45% of the image signal.

The Point Spread Function (PSF) for the visible chief ray (best geometric image) is shown in a 12-micron pixel size sample [v12b]. The same image is shown in a 24-micron square sample [v24b] so it can be compared with the worst case image PSF 1/2-degree off axis [v24w].

Spot diagrams - The spot diagrams are shown for the 12 field positions used to optimize the VIS band and Telescope design. A detector grid is shown superimposed on the spot diagrams. The grid is divided into 12 (V=VIS) and 24 micron (S=SW) wide boxes. The maximum grid square is 48 microns on an edge (M=MW & L=LW). The ZEMAX default unit size is 40-microns for VIS & SW and 100-microns for MW & LW.

Modulation Transfer Function (MTF) Specifications, Goals, and System Performance

The MTF specifications for the AGS Imager are the same as the GOES IJK/LM fractional specifications applied against the Nyquist Frequency Limits. For GOES IJK/LM the limits were for 1- and 4-km.
For the AGS Imager the limits are 1/2-, 1-, and 2-km resolution. The AGS MTF goals are these same limits applied to the higher resolution pixels of 1/3-, 2/3-, and 4/3-km pixels.

The data plotted in the following graphs is the same type for each band, consisting of:

All of the plots are scaled so that X-axis goes from 0 to the Nyquist Frequency (the reciprocal of 2-times the pixel width).

[VNIR] The image performance in the VNIR Band is the most difficult for the optical system. The Telescope was optimized for the VNIR Band. The pixel size for this band is 1/3-km at nadir, which corresponds to an 0.012-mm pixel. The Imager System performs better than the System Goals for the VNIR Band.

[SWIR] The SWIR Band performance just meets the MTF Goals for a perfect, Diffraction Limited System. The real system falls below the goals by a few percentage points at the specification frequencies. On the whole, this is probably good enough performance for the 2/3-km pixels, but there are ways to improve the performance. See the discussion at the end of the MWIR MWIR charts just below.

[MWIR-1] The MWIR-1 Band performance just meets the MTF Goals for a perfect, Diffraction Limited System. The real system falls below the goals by a few percentage points at the specification frequencies. On the whole, this is probably good enough performance for the 4/3-km pixels, but there are ways to improve the performance. See the discussion at the end of the MWIR-2 chart just below.

[MWIR-2] The MWIR-2 Band performs better than the System Goals at the specification frequencies. There is only one difference between MWIR-1 and MWIR-2; the integration time for each pixel. MWIR-1 integration time is equal to the TDI interval and decreases the MTF in the in-scan direction. The integration time for MWIR-2 is much less than the TDI interval and there is a significant increase in MTF performance with respect to the GOAL values. This improvement suggests several factors which might improve the SWIR and MWIR-1 performance.

There are 3 options (not including increasing the telescope diameter) which might be considered:

[LWIR 1 & 2] The LWIR 1 & 2 Bands have different integration times, but the absolute value of the integration times is small relative to the TDI time. The differences in performance caused by the different integration times is not detectable in a graphical presentation, so the LWIR-2 data was not plotted. The LWIR System Goals were reduced because the system is diffraction limited and the only way to meet the goals that were applied to the shorter wavelength bands is to build a larger diameter telescope that results in a heavier instrument and significantly more expensive mission. The larger telescope was not chosen for the GOES system and it is assumed that it will not be chosen for the AGS Imager likewise. The optical performance of the LWIR Bands meets the specified goals.

The resampling procedure is complex (an understatement) and has not been selected yet. Two resample convolution examples have been considered (follow the link above for more details) in order to get an idea how the resampled data might behave:

The LANDSAT interpolation was used for the following calculation because it produced the lowest performance, but also exceeded the performance specification. It was also used because the convolution was developed by experience and preferred when a more efficient MTF conversion might have been selected. There are more problems to conversion than MTF, for instance, aliasing.

pixel size	cycles/radian	MTF specification	cycles/mm	AGS Imager Optical System MTF	Resampled MTF (minimum)	Resampled MTF (mean)
0.5-km VNIR	9000 18000 27000 36000	0.92 0.73 0.53 0.32	7.0 14.0 21.0 28.0	0.96 0.88 0.78 0.66	0.96 0.88 0.76 0.45	0.96 0.88 0.77 0.55
1.0-km SWIR	4500 9000 13500 18000	0.92 0.73 0.53 0.32	3.5 7.0 10.5 14.0	0.98 0.91 0.70 0.56	0.98 0.88 0.62 0.38	0.98 0.90 0.66 0.53
2.0-km (MW-1)	4500 9000 13500 18000	0.92 0.73 0.53 0.32	1.75 3.5 5.25 7.0	0.93 0.84 0.71 0.57	0.93 0.82 0.62 0.39	0.93 0.83 0.67 0.48
2.0-km (MW-2)	4500 9000 13500 18000	0.92 0.73 0.53 0.32	1.75 3.5 5.25 7.0	0.95 0.88 0.79 0.68	0.95 0.85 0.70 0.46	0.95 0.87 0.74 0.64
2.0-km (LW)	4500 9000 13500 18000	0.86 0.63 0.42 0.25	1.75 3.5 5.25 7.0	0.89 0.77 0.64 0.51	0.89 0.75 0.56 0.34	0.89 0.76 0.60 0.43

AGS-Imager Optical/Detector Performance superior to GOES R Specification - The data above shows the Imager Instrument MTF performance exceeds the GOES "Super-R" performance specification at 1/2, 1, and 2-km. The reason for the AGS improved performance is that the detectors allow an area resolution nearly an order of magnitude better than GOES. With other advantages (no central obscuration, rigid cross-scan detector alignment, etc.) the performance is really an order of magnitude better. This improvement does not come free of charge - the data rate increases by more than the order of magnitude and the overall instrument size increases.

pixel size	cycles/radian	MTF specification	cycles/mm	MTF computed
1/3-km VNIR	13420 26840 40260 53680	0.92 0.73 0.53 0.32	10.4 20.8 31.3 41.7	~0.93 ~0.78 ~0.61 ~0.42
2/3-km SWIR	6710 13420 20130 26840	0.92 0.73 0.53 0.32	5.2 10.4 15.6 20.8	~0.90 ~0.70 ~0.50 ~0.30
4/3-km (MW-1)	3355 6710 10065 13420	0.92 0.73 0.53 0.32	2.6 5.2 7.8 10.4	~0.89 ~0.71 ~0.50 ~0.30
4/3-km (MW-2)	3355 6710 10065 13420	0.92 0.73 0.53 0.32	2.6 5.2 7.8 10.4	~0.92 ~0.79 ~0.64 ~0.47
4/3-km (LW)	3355 6710 10065 13420	0.86 0.63 0.42 0.25	2.6 5.2 7.8 10.4	~0.83 ~0.64 ~0.45 ~0.28

The performance at 1/3-, 2/3-, and 4/3-km resolution does not meet the goals set forth in all cases, but it comes close enough. When this level of performance data is resampled, the resampled performance is better than the required performance for 1/2-, 1-, and 2-km pixel mapping.

Although not meeting goal performance, the system does meet specified performance. If it becomes necessary to improve the as-scanned performance for the SWIR and MWIR, the in-scan MTF can be improved by increasing the TDI pixel count and decreasing the integration time for SWIR and MWIR-1.

The maximum (center to corner) grid distortion is about 2 percent. The distortion is slightly greater in the long direction of the detector arrays than across the narrow width. There is negligible (about 0.2 percent) curvature of the grid lines in the scan direction. The data presented is for the VIS band and should be virtually identical for all bands. In the displayed image, a perfectly square grid of source points is traced to the image plane. The displacement from a perfect grid is shown by a plotted "X". There is a line drawn from the reference point to the center of the "X", which is the grid distortion.

Tolerance analysis has been completed in detail for the Visible Band, the most sensitive to alignment because of the small detector size, 0.012-mm square, and images that are dominated by the geometric character of the optics, not diffraction. For all the other bands, the performance at the Nyquist frequency has been analyzed.

[c] VIS Through Focus MTF - The through focus MTF for the central ray shows the system tolerance to defocus. The MTF is for the Nyquist limiting frequency which corresponds to 2 times the pixel width (the highest frequency sine wave that can be reconstructed using 12-micron square pixels). The peak MTF value is about 0.74. The performance specification for this frequency is 0.32. For analysis, the range which allows degradation from 0.74 to 0.60 is considered a reasonably allowable range for optical tolerance. The remainder of the MTF will be used up by the detector pixel size and the dynamics of image collection and display. Note that this analysis is for the 1/3-km pixel size, not the 1/2-km size required by the specification. This difference provides extra performance margin. A reasonably deep defocus is allowed (about 60 to 80-microns, TTHI=0.030-0.040) with minimal degradation in MTF performance. Careful focus alignment taking advantage of tangential and sagittal characteristics can increase that defocus depth to about 90 to 100-microns.

Overall, the tolerance sensitivity is slightly improved over the GATES afocal telescope with transmission optical reimaging because the AGS telescope is slightly slower than the GATES and the GATES field of view was double the AGS field.

Analysis at Nyquist frequency for VIS Band
System MTF	Detector MTF	Optical MTF	Merit Function	Equivalent Defocus (mm) + or -
0.42	0.57	0.73	0.0038	0.000
0.37	0.57	0.65	0.005	0.035
0.32	0.57	0.57	0.006	0.050

Analysis at Nyquist frequency for SWIR Band
System MTF	Detector MTF	Optical MTF	Merit Function	Equivalent Defocus (mm) + or -
0.30	0.41	0.73	0.0077	0.000
0.29	0.41	0.71	0.0083	0.035
0.29	0.41	0.70	0.0089	0.050

Analysis at Nyquist frequency for MWIR-1 Channels
System MTF	Detector MTF	Optical MTF	Merit Function	Equivalent Defocus (mm) + or -
0.30	0.41	0.75	0.0071	0.000
0.30	0.41	0.75	0.0077	0.035
0.30	0.41	0.75	0.0084	0.050

Analysis at Nyquist frequency for MWIR-2 Channels
System MTF	Detector MTF	Optical MTF	Merit Function	Equivalent Defocus (mm) + or -
0.47	0.63	0.77	0.0071	0.000
0.47	0.63	0.75	0.0077	0.035
0.47	0.63	0.75	0.084	0.050

Analysis at Nyquist frequency for LWIR Band
System MTF	Detector MTF	Optical MTF	Merit Function	Equivalent Defocus (mm) + or -
0.28	0.64	0.44	0.0086	0.000
0.28	0.64	0.44	0.0092	0.035
0.28	0.64	0.43	0.0098	0.050

All the bands were analyzed for Merit Function decreases equivalent to defocus displacements of the Focal Plane Assemblies. The VIS band was tested for degradation which would bring the System Performance down to just meeting the Goal Performance. This is a conservative analysis because the resampled data will still be better than the specified performance at 1/2-km grid spacing. The VIS and SWIR Focal Plane Assemblies are physically mounted together and the MWIR and LWIR Focal Plane Assemblies are physically mounted to the same cold finger. The VIS equivalent defocus was applied to all Bands because they are all mounted to the same optical bench in the same style. Reviewing the Resampling MTF Performance, these defocus equivalents could be expanded further and the resampled data would still meet requirements. Thus, the AGS Imager system seems to have adequate tolerance margin for real performance.

The Handbook of Optics gives various sets of tolerance values and categorizes them in increasing difficulty to be achieved as:

The tolerance values applicable to the AGS Imager lie in the range between Commercial Production and Precision Production. That is not to say that the achievement of these tolerances will be easy, but that they can be readily achieved if design and selection of structural materials is carefully considered. The way that ZEMAX studies Monte Carlo tolerance performance is to assign each optical element a position or value randomly selected within the assigned tolerance limits. When this was done with the limits finally selected for the AGS Imager, ZEMAX predicted that a "production" Imager could be assembled within tolerance about 10% of the time. That is to say that a fully assembled Imager would be found to be in acceptable focus the first time the focus of the full assembly was tested. In actual assembly the system will be tested and readjusted until it is well within the allowed limits. In some cases allowances might be made for known thermal effects between laboratory and flight conditions, as they have on other flight instruments. All-in-all, the analysis shows that this is a precision instrument, but not exceedingly difficult to assemble or align for personnel that are used to space flight hardware.

From the Scan Mirror to the Tertiary, the locations are on the central ray. For the subsequent surfaces, the point is located on the surface at the approximate center of the optical element.

VIS-F04 is not an optical match to the 3-D model which was folded in AutoCAD. It is a "sensible" approximation for tolerance sensitivity analysis only.
TRAD: Is equivalent to a defocus (TTHI) which can be compensated for in alignment.
- : not applicable to tolerancing
n/s: not sensitive over range tested
xy : x & y axes aligned with surface, not sensitive. See Thickness component

The following table is a list typical reflectance or transmittance values used to calculate the system transmittance. The values are the ones used by the focal plane engineers to derive the performance and performance requirements of the focal plane arrays. The system transmittance is the product of the transmittances or reflectivities of all the optical elements. These values are neither overly optimistic or conservative. The focal plane arrays are not marginal in their predicted performance based on these system transmittance estimates. The focal plane arrays can be electronically adjusted to compensate for changes in system transmittance, whether improved (higher transmittance) or degraded (lower transmittance).

The Merit Function of 0.0039 is for the very best optimized focus conditions. These data points were derived by manually changing the location (TTHI, thickness) of the Focal Plane Assembly. The Defocus is a plus or minus range from best focus. As the system is defocused, the Mean Radii of all the test locations clusters more tightly. The radius of 9 microns is the system specification for 0.5-km resolution, which is why the 0.005 Merit Function was chosen a a limit for the optical system.

ZEMAX Tolerance Runs were executed for the following tolerance ranges applied to each ZEMAX generated variable. ZEMAX works in local coordinates and does not generate variables that will have no significance, e.g. translations within a plane and tilts normal to a plane are omitted. ZEMAX computes the merit function for each parameter with the range spread over the extreme tolerance limits specified. The top 110 worst case offenders are listed in order of decreasing change in the Merit Function for each Run. The results were inspected for each parameter and the preceding sensitivity table was generated.

Run ID	(mm)	(degree)	(fringes)
VIS-F04-00	0.00625	0.000125	0.125
VIS-F04-01	0.0125	0.00025	0.25
VIS-F04-02	0.025	0.0005	0.5
VIS-F04-03	0.050	0.001	1.0	ZEMAX Default
VIS-F04-04	0.100	0.002	2.0
VIS-F04-05	0.200	0.004	4.0
VIS-F04-06	0.400	0.008	8.0
VIS-F04-07	0.800	0.016	16.0
VIS-F04-08	1.600	0.032	32.0
VIS-F04-09	3.200	0.064	64.0

Surface or Object	Maximum Concentration (solar constants)	Spot Diameter (cm)	Image Motion for 1-degree/sec scan rate (cm/sec)	Dwell Time on Object (sec)	Off-Axis Sun Angle at "first light" (SA, degrees)
Front Baffles	1	42.5 x cos SA (2)	0 – 2 0 - 2	43200 (1) 20160 (1)	90 (no SS) 42 (with SS)
Primary Mirror (efl=80.6cm)	1	0 - 30	0 - 2	0 - 9128 (1) 0 - 7680 (1)	19 (no SS) 16 (with SS)
Back of 1^st Barrel Baffle (28 cm from Prime Focus)	9	10	1 (when present)	0-10(one point) 0-20(any point)	10 (not every scan)
Secondary Baffle (12 cm from Prime Focus)	41	4.7	4.7	3.9 (one point) 7.8 (any point)	3.9
Secondary Mirror (12 cm from Prime Focus)	41	4.7	4.7	3.9 (one point) 7.8 (any point)	3.9
Field Stop (1-deg=22.7-mm)	699	1.135	2.27	0.5 (one point) 1.5 (any point)	1.25
First Beam Splitter	13 (3)	7x10 (ellipse)	7	0.5 (one point) 1.5 (any point)	1.25
Focal Plane (2.23-cm diameter)	723 (3)	1.115	2.23	0.5 (one point) 1.5 (any point)	1.25

[SEC] The view of the Secondary Mirror seen from the Primary Mirror. The baffles in the main telescope tube are light gray. Their outline shows the circular edge for the main beam from the Scan Mirror to the Primary and the decreasing size openings as the light is concentrated onto the Secondary. The baffle around the Secondary that will be illuminated by the Sun has an outer diameter of about 100-mm and an inner diameter of 55-mm, the size of the secondary mirror.

[FS] This is a view from the center of the Secondary Mirror, looking at the Field Stop. The Field stop is a 1-degree diameter circle. The Tertiary Mirror (green) is seen through the Field Stop. The Field Stop Baffle (yellow) is about 55-mm high and 200-mm wide. The Field Stop diameter is about 23-mm. The solar image will be about 11.5-mm. The baffle will be polished aluminum about 2-mm thick.

The mirror material for the AGS Imager must be selected taking into account specific observing concerns:

The ideal mirror material for the instrument environment would be highly conductive, light weight, and have low thermal expansion.

Conventional low-expansion mirror materials like fused silica, ULE, Zerodur also have low thermal conductivity. Their stiffness is also low. Mirrors based on these materials are generally the heaviest. The weight of mirrors fabricated from these materials is the chief reason for not selecting them. Also, the low thermal conductivity is a problem, but has been addressed in the redesign of the GOES secondary mirror mounts.

Alternative materials like aluminum, beryllium, and silicon carbide are attractive because they are potentially much lighter in weight because they are stiffer materials. Sample mirrors deposited on fiber composite structures look attractive for future development, but have not been fabricated to good enough precision to be considered for visible imaging. Many mirrors have been made from aluminum, beryllium, and silicon carbide. There are numerous manufacturers of both the basic materials and finished optical components. Aluminum and beryllium have been used for numerous spaceflight instruments. Silicon carbide as a mirror material has less history, but several flight quality mirrors and silicon carbide structures have been fabricated and tested by several different manufacturers. The tests include thermal-vacuum and flight qualification vibration testing.

Aluminum is not selected because beryllium is a flight qualified material that results in lower weight mirrors, and the lowest weight practical needs to be selected.

Beryllium has been used to fabricate the GOES Scan Mirrors. The thermal distortions of the GOES 8 Scan Mirror significantly degraded the optical performance for several hours each day, as the sun illuminates the Scan Mirror near local midnight. The lightweight rib structure was redesigned for the GOES 9 and resulted in better performance, but still could be improved. Beryllium is stiffer than aluminum and has a coefficient of thermal expansion about half that of aluminum. Beryllium will likely need a nickel surface coating to provide a low scatter surface, which is a potential for causing thermal deformation problems.

Very lightweight silicon carbide mirrors have been fabricated representative of a GOES-like instrument. Silicon carbide has a coefficient of thermal expansion that is about 10 times lower than aluminum and 5 times lower than beryllium. Manufacturing techniques allow "near net shape" designs which result in light weight mirrors. For the Scan Mirror example, the face sheet was only 1.5 millimeters thick. The mirror met its design goals, but would need to be slightly thicker to be stiffer and more conductive for the AGS goals. Mirrors to 60-cm diameter have been fabricated, well larger than any needed for the AGS Imager. The egg-crate style construction mirrors reported in the literature had optical quality good enough for the visible bands. The monolithic mirrors appeared to have been successfully designed so close to limiting performance that they were the lightest weight, but did not meet the AGS visible requirements (they did meet their IR design requirements). It looks like stiffer monolithic designs or lighter weight egg-crate designs will meet the AGS visible performance requirements. Additionally, silicon optical surface coating provides for a very smooth, low-scatter surface finish on silicon carbide blanks.

Silicon carbide looks like the best choice for the AGS Imager for the Scan Mirror. For similar reasons it should be a better choice for the Primary and Secondary Mirrors as well. Beryllium is a good back-up choice for these mirrors, but there is no "off-the-shelf" design that can be adapted directly to the AGS Imager. Even if beryllium is used, minute attentions to details of lightweighting and mounting will be needed.

The optical surface is 5-mm thick. All other surfaces and internal webs are 4-mm thick. The outer edge is 3-mm thick. There are vent holes for each cell and three larger mounting holes. The following images are a back view with the mirror rotated 45-degrees. Front and side views show no significant detail.

The optical surface is a 5-mm thick shell. A 1-cm wide mounting ring is on the back, centered so 50% of the area of the mirror is inside and outside the radius of the center of the ring. The details of this version of the mirror are patterned after mirrors manufactured by SSG, Inc. under a Phase-2 SBIR; reported in SPIE Volume 2543. The following images are a back view with the mirror rotated 45-degrees. Front and side views show no significant detail.

The Tertiary Mirror looks similar to the Primary under both construction styles. The Secondary Mirror is small and essentially monolithic. It may actually be increased in size to give it more heat capacity.

A three-D AutoCAD model of a silicon carbide Scan Mirror has been created in an SSG-like style.

The views above are isometric hidden-line [HL] and photo-realistic [PR] renditions showing the central strong back mount, the elliptical edge, and 5 stiffening webs. The faceplate, webs, and edges are all 1.5-mm thick. The distance from the face of the mirror to the back edge of the webs is 36.8-mm.

Notes:
[1] Best fit (3 point) sphere at blank center, not optical prescription radius of curvature.

[2] SiC Density: 3.217 (CRC Handbook, 1955); Reaction Bonded, 2.93; HP & Pressure Cast, 3.18 (gm/cc)

	Interpolation Offset Fraction
Fraction of Nyquist Frequency	A=0	A=0.125	A=0.25	A=0.375	A=0.5	Mean MTF
0.00	1	1	1	1	1	1
0.25	1	1	1	1	1	1
0.50	1	1	.99	.98	.97	0.99
0.75	1	0.97	0.93	0.90	0.88	0.94
1.00	1	0.95	0.85	0.74	0.68	0.84

X(mm)	Y(mm)	Z(mm)	Name	Surface #
157.2813	0.0000	0.0000	Scan Mirror	M
0.0000	0.0000	0.0000	Origin (ZEMAX)	6
-681.7993	0.0000	0.0000	Primary	11
33.9643	-282.8445	0.0000	Secondary	13
-96.7997	-300.5687	0.0000	Field Stop	X
-708.8929	-383.5341	0.0000	Tertiary	17
-313.1444	-438.2972	-5.3022	BS VS/ML FS	M, 19
-310.3125	-470.9564	135.3341	Fold Flat VS	M, 22
-296.1063	-471.7198	134.3168	WINDOW VS FS	26
-232.6735	-486.4126	135.3774	BS V/S FS	M, 33
-232.2985	-465.4401	133.8636	VIS FPA W-FS	36
-210.4044	-491.9668	135.7784	SWIR FPA W-FS	-
-268.6839	-441.9653	0.0000	WINDOW ML-FS	-
-93.4992	-468.2993	-1.7882	BS M/L FS	-
-94.7858	-475.3118	-23.7545	MWIR FPA W-FS	-
-69.4082	-473.5390	-0.4514	LWIR FPA W-FS	-

				Volume (cu-cm)	Volume (cu-cm)	Mass [2] (kg)	Mass [2] (kg)
Element	Blank Diameter (mm)	Blank Surface Radius [1] (mm)	Shape	Egg-crate style	SSG style	Egg-crate style	SSG style
Scan	508 x 305	flat	flat	xxx	716	xxx	2.302
Primary	360	1864	Concave	1693	936	5.446	3.011
Secondary	55	571	Convex	xxx	30.4	xxx	0.098
Tertiary	240	638	Concave	765	515	2.461	1.657

Name	Surf #	TEDX mm	TEDY mm	TTHI mm	TRAD mm	TETX deg	TETY deg	TETZ deg	TIRX mm	TIRY mm
Scan Mirror	M	n/s	n/s	n/s	n/s	n/s	n/s	n/s	n/s	n/s
Origin(ZEMAX)	6	-	-	-	-	-	-	-	-	-
Primary	11	0.100	0.037	0.400	0.800	0.002	0.004	n/s	0.800	n/s
Secondary	13	0.200	0.050	0.400	0.800	0.004	0.032	n/s	n/s	n/s
Field Stop	X	-	-	-	-	-	-	-	-	-
Tertiary	17	0.100	0.100	0.400	0.400	0.008	0.008	n/s	n/s	n/s
BS VS/ML FS	(M) 19	xy	xy	0.100	-	0.016	0.016	n/s	n/s	n/s
Fold Flat-VS	(M) 22	xy	xy	0.050	-	0.016	0.016	n/s	n/s	n/s
WINDOW VS FS	26	n/s	n/s	1.6	n/s	n/s	n/s	n/s	0.800	0.400
BS V/S FS	(M) 33	xy	xy	0.050	-	0.016	0.016	n/s	1.6	1.6
VIS FPA W-FS	36	n/s	n/s	0.800	n/s	n/s	n/s	n/s	0.200	0.200
VIS FPA	image	n/s	n/s	0.035	n/s	n/s	n/s	n/s	n/s	n/s

	Optical System Transmittance
Band	VIS	SWIR n	SWIR w	MWIR	LWIR
Wavelength (microns)	0.55	2	2	4	10
Element	Element Transmittance or Reflectivity
Scan Mirror	0.90	0.90	0.90	0.90	0.90
Primary Mirror	0.95	0.95	0.95	0.95	0.95
Secondary Mirror	0.95	0.95	0.95	0.95	0.95
Tertiary Mirror	0.95	0.95	0.95	0.95	0.95
B/S VS/ML	0.75	0.75	0.75	0.75	0.75
Fold Mirror	0.95	0.95	0.95	none	none
Window	0.95	0.95	0.95	0.95	0.95
B/S V/S or M/L	0.75	0.75	0.75	0.75	0.75
Filter, Narrow	n/a	0.50	n/a	n/a	n/a
Filter, Wide	n/a	n/a	0.80	n/a	n/a
Filter Typical	0.60	n/a	n/a	0.80	0.80
System Transmittance	0.235	0.196	0.313	0.330	0.330

Merit Function	Defocus, TTHI (mm)	Mean Radius 80% Encircled (micron)
0.0039	0.000	5
0.005	0.035	7
0.006	0.048	8
0.007	0.062	9